Deep Domain Adaptation

Tensorflow implementation of deep learning based domain adaptation models. See more in implemented algorithm.

Work in progress: Experiment results will be appeared soon!

Get started

git clone https://github.com/asahi417/DeepDomainAdaptation
cd DeepDomainAdaptation
pip install .

install error
You might have install error if your environment dosen't have numpy and cython, due to the pot library. Then, import them before install this repo.

pip install numpy
pip install cython
pip install .

Script

`bin/script_tfrecord.py`

This script converts dataset to tfrecord format.

usage: script_tfrecord.py [-h] --data [DATA]

optional arguments:
  -h, --help     show this help message and exit
  --data [DATA]  dataset name in dict_keys(['mnist', 'svhn'])

`bin/script_train.py`

This script is to train models.

usage: script_train.py [-h] -m [MODEL] -e [EPOCH] [-v [VERSION]]

optional arguments:
  -h, --help            show this help message and exit
  -m [MODEL], --model [MODEL]
                        Model name in dict_keys(['dann', 'deep_jdot'])
  -e [EPOCH], --epoch [EPOCH]
                        Epoch
  -v [VERSION], --version [VERSION]
                        Checkpoint version if train from existing checkpoint

List of Models

Domain Adversarial Neural Network
- Ganin, Yaroslav, et al. "Domain-adversarial training of neural networks." The Journal of Machine Learning Research 17.1 (2016): 2096-2030.
- implementation
Deep Joint Distribution Optimal Transport

heqishan/DeepDomainAdaptation